Renee DeSilva  00:00

Welcome back to the table. I’m Renee DeSilva, CEO of the Health Management Academy and your host. This week, I had the pleasure of speaking to Diana Nole, executive vice president and general manager of Nuance Communications, which is now a Microsoft company. The philosophy of strength in numbers was entwined in our conversation as Diana began by sharing the strategic and complementary partnership of Microsoft and nuance. Rooted in the expansive health and life sciences industry. Their shared vision is focusing on solving complex challenges, with an understanding that no one company can solve all pain points alone. Together, they offer breadth horizontally and scale paired with vertical depth and expertise. Here are a few of my takeaways from our conversation. Today, AI is everywhere. But how would we have defined that just 10 years ago, Diana shared that while AI was around, it was comparable to an early stage of human life, its evolution across the past five to 10 years is nothing short of remarkable. Fast forward to today, and everywhere you turn, you’ll hear the captivation around Gen AI, the fastest growing consumer software application in history. It’s scalable human likeness, this threefold, Insight extraction, user interaction and the broad impact it has on society. Next, Diana’s passion for problem solving and curiosity for technology was evident from an early age. And that led to her attaining dual degrees in computer science and math. She believes that staying curious and acknowledging that technology is leveraged everywhere gave her the competence to be part of important conversations. She’s a champion for diversity, equity and inclusion. With a personal mission to continue inspiring girls and their formative years to pursue math, science and tech fields. We concluded our conversation where we began discussing what makes a good partner Diana shared the hallmarks of a good partnership include finding a win win for everyone involved, understanding what it’s beneficial to both parties, and being honest about who brings in what capabilities. And as Diana described, this is a delicate process of finding equilibrium for all stakeholders. So with that, let’s head to the table. Good morning, Diana, welcome to the table.

 

02:17

Good morning.

 

Renee DeSilva  02:18

Very happy to have you a lot to cover. So I’m going to just go ahead and jump in. So in 2022, Microsoft finalized the strategic acquisition of Nuance Communications. Tell us a little bit more about the why behind that. What’s the shared vision,

 

02:36

the shared vision of Microsoft and nuance is really grounded in the fact that health and life sciences is a very large industry, very complex challenges, no silver bullet, and no one company regardless of what their capabilities are, could solve the various pain points alone. And so Microsoft and nuance share a belief in purpose built technology, not technology for technology’s sake, but a real focus on outcomes, and AI pointed at solving specific problems. Nuance is a trusted brand in healthcare. Our technology is deeply embedded into clinical workflows that hundreds of clinicians depend on across the globe. But this is brought together with Microsoft’s horizontal technology and AI capabilities. They have industry leading security and compliance. And they have the ability to deliver global cloud services, and a partner and developer ecosystem at scale, in support a broad digital transformation. These things are important to our clients, both on the clinical side and on the infrastructure side. So together, we offer the horizontal breadth and scale, along with the vertical depth and expertise needed to create integrated healthcare experiences, that are really focused on outcomes, providing greater access to care, enhancing the clinician patient experience, and hopefully ultimately improving health outcomes.

 

Renee DeSilva  04:14

I love that. That the power of both horizontal capabilities and vertical depth and I think the point that really resonates too, it’s just that there’s really two challenges to serve here, if you will, both the caregiver experience and then taking out some of the maybe friction associated with with with how patients experience their care. Absolutely,

 

04:38

absolutely. As you know, as again, as we’ve learned, no single company can be great at all of these things. So you want to grill these relationships. In this case, it was an acquisition, but also we need to continue to extend these relationships to other partners and other people, other companies that work in the ecosystem because how Healthcare has again, I would probably say at multiple times has the most complex challenges to solve. I

 

Renee DeSilva  05:06

agree, agree. So you have been an executive in technology roles for your career. And I’m going to come back to just sort of how your career has unfolded. But I’d love you’re just zooming out take before we go there around. If you were to have looked back 10 years ago, how would you have thought about the potential for AI compared to what we are experiencing today?

 

05:33

Well, just 10 years ago, you know, AI was around, but it couldn’t reliably provide language or image recognition at a level that was really equal to what I like to refer to as a human adult, where you have to take in complexity and reasoning, I would have defined it as artificial intelligence that was good, but very basic. And again, I would probably go back to it really was representative of the early stages of a human’s life similar to preschool, early school, right, you’re talking about, the computer can identify things it can can, it can identify, and do basic things such as this is a dog, this is a cat, this is bone, this is soft material, etc. But obviously, I have enjoyed the evolution we’re in right now. They AI systems have become steadily much more capable over the last five to 10 years. For example, some implementations of AI systems are available on the phone in your pocket, right, you can get image recognition that can categorize your photos, you can use speech recognition that transcribes what you dictate. And more recently, AI systems are now even beating humans and standardized tests in all domains. Very impressive. Early systems focused on generating images of faces, newer models have capabilities to do text to image generation, based on almost any prompt, and within seconds, that’s really the evolution and you can now describe what you want in the image. And the AI will actually create it with a high level of accuracy. So the level of learning is, you know, is exponential. Then when you think about it, in terms of not only image generating AI, but systems around parsing and responding to human language, right. So emails, get auto completed massive amounts of online texts can get translated, reports can get auto generated, media outlets can publish AI generated journalism, knits really amazing in terms of where we are. So AI is definitely here. These rapid advances in AI capabilities have made it possible to use AI and a wide range of new domains. And obviously, generative AI is faster, more scalable, and approaching human likeness, it’s uniquely able to handle what I would say is three key areas. Insight extraction, rapidly searching large amounts of text, visuals, and identifying relative relevant patterns. It can do content generation, it can develop complex data tailored to specific contexts. And it can do that in text, visual sound. And then it can do user interaction that’s really almost like out of the box human like conversational ability. It’s really no longer a technology of the future. It’s already impacting all of us. I would kind of close and say, though, that at the same time, the performance of AI is still mixed, especially when the real world use case is complex in the level of knowledge and reasoning required. This is where systems still perform much worse than humans. This is where within healthcare, we’re really moving around using generative AI around those things I just mentioned, things that can reduce the administrative burden, not getting into clinical diagnosis. It’s obviously a very general technology, it can be used for extremely good goals. But it obviously can also be used from extraordinarily bad ones. So it’s important that all of us develop an understanding of what is happening and how we want the technology to be used. Very exciting time very optimistic about all of this with the caution that it really needs to serve the right use cases ultimately,

 

Renee DeSilva  09:48

that’s right. And the thing that strikes me about all that you just said was it’s almost like objects appear closer. Objects in the mirror are closer than they appear. I think what everyone is struck by, at least when we have our conversation shins in our in our convening events across the health management academy. It’s just the pace at which the progress has been made, I think is what really strikes folks these days.

 

10:11

Yes, yes, the pace at which and people do talk about like, why now? And certainly everything seems to have converged. We have talent that really understands all of those we have computing power that understands that we’re beginning I think, to get to that level of trust that people have, and that responsibility of what do you use it for? How do you serve up and make sure that whoever the end user is this receiving something that could be generated by AI knows that maybe AI was involved? So the evolution has not just been about the technology evolution, but the evolution of the change in the adoption and the willingness to trust it? Indeed,

 

Renee DeSilva  10:53

so let’s go a little bit more, we were just zooming out, let’s zoom in a bit and talk about you and sort of how your career path has unfolded. You have dual degrees in both computer sciences and math. And so is it fair to assume that early on you had a vision for your career path?

 

11:11

Well, you’re probably giving me more credit than I deserve in terms of the vision for my career has definitely been an evolution, I would say first and foremost, I’m I’m part of a first generation college family, it was extremely important to my parents that we all went to college and, and they really helped us to achieve that, while also generating great values of each of us chose our own path, we work hard and really give back I do. I do love math, I do love solving problems. And so computer science did seem to be a natural selection. Along with it, I would say this problem solving passion has stayed with me throughout my career, I’ve always had a great curiosity about technology for the purpose of generating improvement in business, it served me well. Even though I only did software programming for a very short time after I graduated, I learned early on in my career, that technology is used or leveraged everywhere. So it gave me the confidence to be able to be in the conversations talking about it. But it’s equally about the process and change management for those specific use cases. So it kind of comes back to well, how are you going to actually solve this problem, I would just close and say I’ve had great mentors and leaders around me both to work for and to work with. And here I am working at an amazing technology giant Microsoft, getting the opportunity to work on solutions that are and hopefully will continue to have a bold impact in the industry. So I’m still learning and growing, which is what all of us want in our career, isn’t

 

Renee DeSilva  12:50

it? Absolutely. Absolutely. Just sort of staying on that thread a little bit. I just track closely, just overall gender representation in the industry. And I think the latest women and technology stat is that overall female representation and check in technology jobs sits firmly at 25%. So some some improvement year over year, but but largely sort of at that at that prompt. When you think about technology and AI and all of that, that that implies, how do we continue to promote and groom and mentor and sponsor, more female leaders in this space, anything come to mind for you?

 

13:36

Well, the important thing is businesses and industries, not just tech always benefit from diverse perspectives, you know, backgrounds and experiences. This is obviously well documented. And my experience has always been that at the end of the day, how can you build a team or an organization or a product that best serves everyone? If you have a group of people who all think the same way, while have the same experiences? The answer is you can’t you’re limited to your own views and perspective. And that’s a disservice to everyone and will likely not get you to the full success that you want. And so in order to promote AI and tech as a field for more females, I think we have to continue to inspire encourage girls early on in their formative years, and really break some of the stereotypes and things that hold them back in terms of being good in math, science, engineering technology. We also need to be sure that when we’re even as adults in conversations that we don’t get hung up on lots of acronyms technology speak that not everybody can participate in. I really try to break that down from occurring. I’m a champion for diversity, equity and inclusion. One of the tangible examples that I have done is I endowed a scholarship fund at my undergraduate alma mater for females majoring in the sciences. It’s used to cover the costs of doing internships, which are critical to the maturity and the experience that college students need to be effective when they join the workforce, I also look for equal representation in my own organization and the ones that I join, and I look for that in non traditional roles for women. And so to that end, in the last two years, is one example, I hired two women as general managers in healthcare in healthcare technology. And that has helped us achieve equal gender representation. And I think it’s just so valuable to our ability to produce solutions that appeal to everyone.

 

Renee DeSilva  15:40

That’s right, especially when you even go back to I mean, just even anchoring it on the core, at least as you know, one of the core elements of nuance in terms of DAX, and taking out the burden on documentation. I think there’s some, there was, I think, a JAMA study that showed that female physicians spend more time documenting than male physicians. And so you would then want to also think through the folks that are designing technology and figuring out the adoption, also, you know, sort of represent that same swath of stakeholder so you can see how there’s real power in having your product teams and your executive teams match the workforce, which, with the healthcare workforce, which has such a high representation of a female leader, so I just I love that connection. Dyana.

 

16:30

That’s exactly right. If you can actually understand and appreciate what the maybe not emotional benefit is around why it is so critically important. It just also gives you just another lens as to why it’s so critically important. We need to do it early in the life of adults, young girls, people that really we want to make it attractive for them to get into these all kinds of fields. And every as we talked about early on, Technology is everywhere. So it needs to be representative everywhere in every industry, not just healthcare. Okay,

 

Renee DeSilva  17:09

so let’s, let’s keep moving. Let’s just spend another minute or two on this gen AI, which, you know, sort of is the the topic du jour launched in late 2022. Just the speed at which the user base has caught up I think there’s more than 100 million users. I may be in the rare minority of folks who have actually not actually tried chat GPT yet. I’m a late adopter. But I’d be curious for it does come up as just nuances technology, in particular, being very far ahead of the curve. Just I guess, what was the the early tea leaves that you read? How did you all think about preparing for the launch of this technology? Just bring that to life a bit, if you could?

 

17:51

Well, the good news is that we were part of Microsoft, which Microsoft had formed a partnership with open AI to collaborate on this development. Correct. But at the same time, the interesting thing is that we also had to kind of come up to speed, you know, Microsoft and open aI had agreed to work together on several key areas, which we get to benefit from, it’s made an investment in it, it’s allowed the company to use Microsoft Azure, its as its preferred cloud platform, and really allowed open AI to tap into the vast resources of Microsoft. At nuance. We were part of the discussions between Microsoft and open AI and we prepared by being very thoughtful in the approach to applying it to healthcare workflows. One of the very first things we did was we briefed our executive client counsel, which is a group of very different types of customers that we serve. So it spanned across all the various segments, whether it’s large hospitals, small outpatient, physician, group, radiology. And we really asked them, How do we see how do you see generative AI manifesting in healthcare, we talked a little bit about our early views and our plans to incorporate it into DAX, we gave them a private preview of Doc’s Express starting in June of this year. So I think that the one thing that’s really at core of what we always do is bring our customers in as quickly as possible, and ask them for their opinions, because they will be clear on where they think it’s to be used, and where it should not be used. And that helps us build out a very responsible approach to generative AI that I think sets us apart in terms of how maybe some organizations are just really rushing to implement it.

 

Renee DeSilva  19:45

And yeah, I love that. And I guess just staying on that that theme is when you think about the way in which we do this, that’s grounded in just good governance principles and the right safety and responsibly just it sounds like you’re getting guidance from the folks Your executive counsel, are there other ways in which you think about that?

 

20:04

Well, we do use this private preview process where a select group of customers tests the solution before it’s released to the entire industry. We also, we always build solutions where a human in this case, the clinician is in the loop, they ultimately sign off on the work that AI generates. And also based on their use of the solution, we then take that learning back end to make the AI better and better, less biased and much more responsible in terms of the guardrails around how AI should be used. We do certainly adhere to Microsoft’s responsible AI standards, which also touches on data stewardship. You know, your data is your data. It’s only used to train and fine tune your AI. And it’s protected by Microsoft’s cloud in terms of compliance and security controls. So it comes at it from a variety of areas in terms of what’s important.

 

Renee DeSilva  21:05

Yes. And so just sort of staying on that your data is your data, your your your owning it, I guess let’s talk through I was sort of struck in a recent conversation, as you know, the Health Management Academy partners with Microsoft and an AI summit and when I and we had about 40 to 50, Health Care Health System leaders who were owning the AI strategy at their organizations, and we talked a lot about some of the headwinds and this sort of unfolding. Well, the regulatory complexity, the org design and talent implications. I’d love your your take on that, you know, how do we as an industry, both on the provider side, and companies in the ecosystem, just appropriately, partner around that?

 

21:52

Well, I think what was great about the summit was the fact that they really were looking at what sort of the internal roadmap to building scalability in terms of how to use AI, they were looking at sequencing and prioritization and culture and policies and governance. And to your point. While there are many things that are sort of tailwinds here, there are also some headwinds to really be responsible about it, in terms of you said, regulatory, you know, what is the use of it? In terms of can we get through interoperability issues? Can we get the accuracy to be of a level that gets maximum adoption? Do we have the infrastructure to engender the trust? Do we have the organizational design and the available talent? This are some of the unlocks that we need to address some of the headwinds that we know are there, so that we can really get it to a point of scalability, which I loved really about the summit, it was all about investing in, where can we get scalability with this? That’s

 

Renee DeSilva  23:01

right. Yeah, I would note too, you know, I think there’s a lot of promising pilots and test cases. And I think there’s just a big appetite for figuring out how that that works at scale. The other thing that I noted was that I believe health systems are being very deliberate about their archetype, right? So defined as, where should they like, what systems are equipped to do this, perhaps more independently? Where are where does there need to be sort of a partner first mantra. And so I, you know, I just I’d love I’d love your take on that. I think that health systems in particular, are taking cues from respected industry partners on what they should take on directly versus where they should differ more broadly. How would you advise in terms of really thinking about making those investments appropriately?

 

23:54

Well, we are in a lot of conversations with customers right now learning from them as much as trying to educate them in terms of where we’re seeing this be used and where we’re not. We’re also offering some of our ongoing and emerging views on these questions. Right now, where I see a lot of investment is really what I would refer to as discovery. Where can the technology help solve real issues now and some proof of concepts I’ve seen, for example, many people have seen sort of the aspects of drafting email responses for patient portals. And what is good about that in terms of 80, sometimes viewing more empathy, and then what is not so good in terms of maybe just ensuring that the patient is aware that this is somewhat generated by AI but signed off by a clinician. So this first phase of where people are really investing is really where it should be, and that is where can technology help solve real issues. Once they have that prioritized use cases that they want to engage John, then what they’re looking at is what’s available right now. So they’ll ask us as vendors, what do you have? What’s available? What are the solutions that we could invest in, that we could be one of your private preview customers, but tell us more about what’s on the roadmap. So at this past year, one of our trade shows we did talk about all along the journey of a patient, where you might see AI infused, and we were very clear about where some of these things might be us providing them might be co developed with them, or might actually be provided by a partner. I think what is also different right now is the level of tools that are available that allow you to do things potentially in house, but to do it with tools that still give you the scalability that you want, so that you’re not producing something that’s just a one off niche type solution, but something that can be reused, repurposed and can grow is really something that also has evolved over the last 10 years.

 

Renee DeSilva  26:14

Yeah, I think that’s right. So then, so then maybe, as a maybe one final question before, we before we round out, when you think about partnerships, broadly, and you mentioned the discovery phase, and really meeting people where they are on that journey and mapping back to maybe mapping back to where they’re trying to go. talk through what you’ve seen throughout your body of work around what makes for a good partner, and particularly from the maybe from the health system lens out like when you work with health systems, I know I’m sure you know that some are really primed and work you work with effectively, like what are the hallmarks of the most successful partnerships? How did those organizations show up? What advice would you give to health systems on how to be a good partner as they’re looking to accelerate this journey?

 

27:15

Well, oftentimes, from a good partnership, the good partnership tries to find a win win situation for everyone and understands what is beneficial to each party. Sometimes what makes a good partnership, the best partnership is people that come to the table and say, I can’t do that, or that’s not in my wheelhouse to do, I don’t have that capability. And so a good partnership will be honest about who brings in what capabilities, and is really focused on speed, and agility and being able to get something out that can actually produce an outcome and a result. If you see things that are taking so long that you’re not going to really be competitive, I think that’s when you can kind of say to yourself, This partnership just doesn’t have the right winning combination. But if you go in and you say, these are the assets and the capabilities that you can bring, and these are the ones that I can bring. And we can do this, and we can do it quickly, and we can iterate. And we both have a winning proposition and we get something out of this. That’s where I’ve seen partnerships work the best. For us. That means sometimes that we have to think not only about doing everything, right till the very last mile, you know, right to the very end of creating the solution, but looking at why would it what would be the best solution for the end user. And oftentimes, that’s that it needs to be in the workflow that they’re used to using. In our case that oftentimes is the electronic health record. And so we look to build great relationships with our electronic health record vendors to really embed our solutions in a way that’s very easy for the end user to use. And that creates better adoption. We’ve seen it over and over it has to be in the workflow. And if you can get the high levels of adoption right out of the gate that really proves that you’ve got a good partnership.

 

Renee DeSilva  29:19

Yeah, I love that finding the equilibrium in terms of where we’re really multiple stakeholders. You mentioned EMR as a as a sort of a third piece of the puzzle, but just where we’re all really able to do our highest and best I think is a great aspiration. Absolutely. So one final question, which does not necessarily need to be connected to the conversation today. But it’s a question that I ask all of my guests, so we’ll land here. But if you could invite any two people to a table that you curated for really any conversation, perhaps it’s continuing on this thread. Maybe it’s just two people whom you’d like to just spend more time with. What who would you invite and why?

 

30:01

I’ve been in a career in healthcare in technology, as you noted, and I’ve had his passion for problem solving, no matter what the problem is, it always can be solved. But the best way to do that, I believe, is to talk to customers, and then bring your engineers along with you. So if I was going to curate a table, it might seem a little boring. But it would be to pick a customer and to pick like one of my, you know, seasoned, but unbiased technology leaders and bring them to the table and say, What’s the biggest thing that we need to work on? And how would we approach it, I would always be motivated, encouraged, just really even more passionate about being able to solve big things. That’s what I really love doing day in and day out. And that’s what I tell my organization, if I can work with us to solve some big challenging issues and do it and have a bold impact in the industry. That to me is fun. It’s hard work, but I enjoy it.

 

Renee DeSilva  31:03

Well, I think the good news on that, Diana is that we are we are ripe with challenges to solve. And I do think our industry is really ready to meet the moment. So no doubt that that will be a conversation that you’ll be able to have. I’ve appreciated chatting with you. And just really thank you for your thoughts and your insights today. And I think this is probably a good place to land.

 

31:24

Well, thank you so much. I’ve really enjoyed the conversation.

 

Renee DeSilva  31:28

Thank you. Thanks for joining me at the table with Renee DeSilva, a podcast brought to you by the Health Management Academy. I hope you enjoyed this episode. And if you did, subscribe, and drop us a review on Apple podcast, Spotify, or wherever you’re listening to this podcast now for all of our episodes, including show notes and transcripts and more information about how you might join me at the table in the future. Please head to H M academy.com/podcast. I look forward to having you back at my table next time. Talk to you again soon.